Dataprefetching Framework for Mobile Behavior Mining Using Co-Smart-Cast Method
Data Mining is a widely used technique for discovering valuable information in a complex data set. In this paper the authors discussed the main issues about Mobile User Behavior Mining. In mobile transaction database, users in the different user groups may have different behaviors. So that they proposed parameter-less clustering algorithm named CO-Smart-CAST. This behavior mining can produce better accuracy of predicting the mobile user information using that algorithm and to provide efficient mining of sequential access in mobile patterns. Here, the clustering method provides a way to calculate the time interval between one transaction to another transaction. They have predicted the behavior of mobile user using COSmart-CAST method.